Validation of a proxy‐reported SARC‐F questionnaire for current and retrospective screening of sarcopenia‐related functional impairments

Background The strength, assistance walking, rise from a chair, climb stairs, and falls (SARC-F) questionnaire is a well-established instrument for screening of sarcopenia and sarcopenia-related functional impairments. As it is based on self-reporting, its use precludes patients who are unable to answer the questionnaire as a consequence of severe acute diseases or cognitive impairment. Therefore, we aimed to validate a proxy-reported version of the SARC-F for both ad-hoc as well as retrospective screening for severe sarcopenia-related functional impairments. Methods Patients aged ≥60 years completed the SARC-F and performed the short physical performance battery (SPPB) at baseline (T1). Proxies in Cohort A gave a simultaneous assessment of the patients' functional status with the proxy-reported SARC-F at T1 and again, retrospectively, after 3 months (T2). Proxies in Cohort B only completed the SARC-F retrospectively at T2. The questionnaires' performances were assessed through sensitivity/specificity analyses and receiver operating characteristic (ROC) curves. For non-inferiority analyses, results of both the patient-reported and proxy-reported SARC-F were correlated with the SPPB total score as well as the results of the chair-rise test subcategory; the respective correlation coefficients were tested against each other. Results One hundred and four patients and 135 proxies participated. Using a SPPB score < 9 points as the reference standard, the proxy-reported SARC-F identified patients at high risk for sarcopenia-related functional impairment with a sensitivity of 0.81 (ad-hoc), 0.88 (retrospective Cohort A), and 0.87 (retrospective Cohort B) as well as a specificity of 0.89 (ad-hoc), 0.78 (retrospective Cohort A), and 0.64 (retrospective Cohort B). Areas under the ROC curves were ≥ 0.9 for the ad-hoc proxy-reported SARC-F and the retrospective proxy-reported SARC-F in both cohorts. The proxy-reported SARC-F showed a non-inferior correlation with the SPPB compared with the patient-reported SARC-F for ad-hoc (P = <0.001) as well as retrospective screening for severe sarcopenia-related functional impairment in both Cohorts A (P = 0.007) and B (P = 0.026). Conclusions Proxy-reported SARC-F is a valid instrument for both ad-hoc as well as retrospective screening for sarcopenia-related functional impairment and could become the standard tool for evaluating this risk in older adults with severe acute disease, for example, in patients with quickly evolving haematological conditions.